A few OS X problems

[mikesmith]

I dusted off my One/500K board today to see where things were up to, and got a bit hopeful about the ZAP IDE; I've wanted for a long time now to play with the ZPU, but it's just never worked out.

 

There are still a few problems - since issues are disabled on the github page this seems like the only place to make a list.

• When attempting to build and upload using the AVR8 core, compilation fails because the IDE attempts to run "cp. exe".
• When attempting to run the flash-the-bootloader pass for the ZPU core, flashing fails because this line in platform.txt:

• tools.papilioprog.erase.pattern="{path}/bin/papilio-prog.exe" -c

• In general there aren't any OS X binaries in hardware/zpuino to substitute here.
• The Papilio QuickStart example uses what seems to be an outdated call to Serial.print:

•   Serial.print(thisByte, BYTE);

  which seems like it needs to be

• Serial.write(thisByte); 

• After flashing zpuino-1.0-PapilioOne-S3E500-Vanilla-1.0.bit using the OS X Papilio Loader GUI (had to temporarily unload the FTDI kext as discussed elsewhere), attempting to build and upload the patched QuickStart example ultimately fails with:

• Cannot get programmer version, aborting

• I tried invoking the uploader manually, including at 115200 baud, with no luck either:

• ./build/macosx/work/Arduino.app/Contents/Resources/Java/hardware/tools/zpu/bin/zpuinoprogrammer -s 115200 -R -d /dev/cu.usbserial-00001214B -b /var/folders/v1/5v5zh56164x0ltx6gpbhyldm0000gn/T/build5681735252102643770.tmp/Papilio_QuickStart.cpp.bin -vvvvvvvvvvOpened device '/dev/cu.usbserial-00001214B'Connecting...[1391668753.446547] Tx: 0x7e 0x01 0x1e 0x0e 0x7e[1391668753.647616] Tx: 0x7e 0x01 0x1e 0x0e 0x7e[1391668753.848701] Tx: 0x7e 0x01 0x1e 0x0e 0x7e

• I built a fresh copy of the programmer from the HDL repo as discussed elsewhere here, with the same results (so I'm wondering how I can tell whether the ZPU is even alive in my FPGA?).

 

[Jack Gassett]

Ok, this is good information, I want to tackle ZAP IDE on OS X soon... I thought Alvie had worked with someone to get all the tools working for OS X already, but I did not include that in the latest release.

 

One thing to note is that the AVR8 is not going to work without a major rewrite. It uses the data2mem tool which is only available on Windows and Linux, we are dead in the water as far as OS X goes.

 

But, ZPUino should be very much doable. It won't be possible to edit schematics since that requires Xilinx ISE but everything else will be possible.

[mikesmith]

Thanks for the heads-up on the AVR8; I guess the only path forward there would be with a bootloader similar to the ZPU. It might be interesting to someone else, but for me I was only interested in getting *something* to work to validate that I had a working board and bootloader.

 

Feel free to ping me if you want me to try something...

[Jack Gassett]

Ok, I just sorted out virtualization on my computer and have OS X running again. I'm syncing ZAP IDE on the Mac now...

[mikesmith]

I tried running through the quickstart on Windows as well, but I have the same issue (timing out waiting for the bootloader)

 

The AVR8 example also fails. If I tell it to write to flash, the program and verify phases all pass, but I don't get any serial output from the QuickStart example. Every now and then there's a burst of activity on the RX LED, but neither the Serial Monitor nor Teraterm can make anything of it.

 

This is all on a v2.05 Papilio One/500K, if that makes any difference...

[Jack Gassett]

It sounds like you might have the wrong com port or the wrong speed... I would recommend plugging in the Papilio board, looking at the serial ports available in ZAP, then unplug the board and verify they go away. Plug the board back in and if they come back you have the right ports. You should be selecting the 2nd serial port that shows up, but if that doesn't work try the first one, sometimes they get turned around... For the speed, look at the speed defined in the setup loop and make sure the ZAP serial monitor is set to the same speed.

 

Let us know how it goes.

 

Thanks,
Jack.

[mikesmith]

Jack, thanks; I should have noted what I'd tried in more detail - I am indeed looking at the second serial port and at the bitrate set in the sketch setup function.

 

Regardless of the port I'm looking at though, I'd expect to be seeing the serial LED blinking since the sketch looks like it should be transmitting regularly.

 

If there's nothing known to be problematic with the v2.04 (typo above, sorry) boards, then is there anything else I can try to establish confidence in the board (or to confirm that it's died)?

[Jack Gassett]

If the problem is the board then we will replace it for you, no questions asked. I would just ask that you try everything you can to make sure it is a problem with the hardware. The best bet is to use a Windows machine to load the Quickstart bit file. If you do not see serial communications having done that then shoot us an email at support@gadgetfactory.net with a link to this thread and a current mailing address.

 

The bottom line is that we want people who are interesting in playing with the Papilio to have a working Papilio, so we will send you a new one if yours got toasted somehow.

 

Thanks!

Jack.

[mikesmith]

Jack, thanks for the offer. The closest I have to a Windows machine is a VM; the bitfile does appear to load and verify correctly.

 

I am more inclined to view this as a fine opportunity to upgrade to the new Papilio, as at least one of my ZPU tinkering plans will require a lot more RAM than the '500 board has.